给出了一种能和网络结构一一对应的、合适的染色体编码方法.用物种入侵的遗传算法训练人工神经网络, 在入侵过程中, 遗传算法自适应地调整交叉算子和变异算子.提出了一种根据平均适应度值确定入侵物种规模的方法, 并详细描述了算法步骤, 最后通过实验证明了本文算法的有效性和优越性.

A suitable chromosome encoding method, which could correspond with the network one by one, was proposed. The species invasion genetic algorithm was used to train artificial neural networks. In the invading process, the genetic algorithm adjusts adaptively crossing operation and mutation operation. A method based on the average fitness values was proposed to determine the scale of invasion species, and a detailed description of the algorithm steps was given. Finally, the validity and superiority of the algorithm are proved by the experimental results.